The Philosophy of Grammar

THE PHILOSOPHY

OF GRAMMAR' BY

OTTO JESPERSEN

LONDOM

GEORGE ALLEN & UNWIN LTD

IlEPRII",FD

FIRST PUBLISHED IN 1924 1925, '929, 1935, '948, 1951, 19S5

",J)

T /", hook u copyrrght under tl" Berne ConventIOn. Apart from any fair dealtng for the purposes of private study, research, en/IC1S'" or revIew as prr-

mlfted under the Copyrtght Act 1956, no portIon may be reproduced by any process without wrlttm permISSIon. EnqUirY should be made to th. ?ubltshe,

© George Allen

& Unwrn Ltd., 1924

PIlINTED IN GREAT BRITAIN BY UNWIN BROTHltRS LTD., WOI:ING AND LONDON

195 8

To my old friend and everwilling helper

G. C. MOORE SMITH

PREFAOE n fau. bl'aucoup de philosophic pour s&voir observer une Coil ce qu'OD voit taus lee jours.-ROU8SEA.U. THIS book has taken long in making, and like other pet children, it has borne many names. When I gave the first crude sketch of it as a series of lectures at Columbia University in 1909-10, I called it an Introduction to English Grammar; in the preface of the second volume of my ~Modern English Grammar (1914) I was rash enough to refer to " a forthcoming book on The Basis of Grammar" ; in Language (1922) I spoke of it again as "a future work, to be called, probably, l'/te Logic of Grammar," and now at last I venture to present it under the perhaps too ambitious title of "The Philosophy of Grammar." It is an attempt at a connected presentation of my views of the general principles of grammar, views at which I ha.ve a.rrived after long years in which I have studied various languages and have been preparing an extensive work on English Grammar, of which I have so far been able to bring out only two volumes. I am firmly convinced that many of the shortcomings of current grammatical theory a.re due to the fact that grammar has been chiefly studied in connexion with ancient languages known only through the medium of writing, and that a correct apprehension of the essential nature of language can only be obtained when the study is based in the first place on direct observation of living speech and only secondarily on written and printed documents. In more than one sense a modern grammarian should be novarum rerum studio8U8.

Though my concern has been primarily with linguistic study, I have ventured here and there to encroach on the territory of logic, and hope that some parts of my work may contain things of interest to logicians ; for instance, the definition of proper names (Ch. IV), the discussion of the relation between substantive and adjective (Chs. V and VII), the definition of 'abstracts' as nexuswords (Ch. X), the rela.tion of subject and predicate (Ch. XI), and the tripartitions in the chapter on Negation (Ch. XXIV). 7

8

THE PHILOSOPHY OF GRAMMAR

I have had many difficulties to contend with in writ.ing this book; one of these is the proper arrangement of my chapters, inasmuch 83 the subjects they deal with interlock and overlap in the most bewildering way. My endeavour has been to avoid as far as possible references to subsequent sections, but it is to be feared that the order in which different topics are presented may here and there appear rather arbitrary. I must also ask the reader's indulgence for my inconsistency in sometimcs indicating and sometimes not indicat.ing the exact place where I have found a passage which I quote as an example of some grammatical phenomenon. This has not been found as necessary here as in my Grammar, where it is my principle to give exact references to all passages quoted; but many of the phenomena mentioned in this volume are such that examples may be easily found in almost. any book written in the language concerned. OTTO JESPERSEN. UNIVERSITY OF COPENHAGEN,

January 1924

Since this book was first published (in 1924) I have carricd out and further developed some of the ideas it contains in volumes 3 and 4 of my Modern English Grammar and in Bssentials of English Grammar to which the reader may therefore be referred. O.J. LUNDEHAVE, HELSINGOR (Er.SINORE).

NOtJllmber 1934

CONTENTS PAOli:

Preface

7

Abbreviations of Book Titles, etc.

11

Phonetic Symbols

15

Living Grammar

17

II. Systematic Grammar

30

CHAPTER

I.

III.

Systematic Grammar (continued)

45

IV.

Parts oj Speech

58

V.

Substantive8 and Adjectives

72

VI.

Parts oj Speech (concluded)

82

The 'Three Ranks

96

VII. VIII. IX. X. XI. XII.

Junction and Nexus

108

Various Kinds of N ex'U/;

117

N exus-Substantive8. on Nexus

Pinal Words

Subject and Predicate Object.

Active and Passive

133 145 157

XIII.

Case

173

XIV.

Number

188

Number (concluded)

202

Person

212

XV. XVI.

9

10

THE PHILOSOPHY OF GRAMMAR CBAl'TEB

PAGB

XVII. Sex and Gender XVIII.

Oompari8on

XIX. Time and Tense XX. Time and Tense (concluded)

226 244 254 269

XXI. Direct and I ndired Speech

290

XXII. Olas8ification of Utterancu

301

XXIII. Moods

313

XXIV. Negation

322

XXV. OoncZU8ion

338

Aprendix

349

Index

353

ABBREVIATIONS OF BOOK TITLES, ETC. Asboth Gramm = O. Asboth, Kurze ffissische Grammatik, Leipzig 1904. Bally LV = eh. Bally, Le Lnngage et la Vie, Genove 1913. St = Traite de Stylistique Fran('aisp., Heidelberg 1909. Bloomfield SL -= L. Bloomfield, An Introduction to the Study of Language, - ......... New York 1914. Boyer at Sporanski M = P. Boyer et N. Speranski, Manuel pour eEtude de la Langue Russe, Paris 1905. Bradley ME = H. Bradley, The Making of Engli.~h, London 1904. BreaJ M = M. Breal, Me/angM de Mytlwlogie et de Linguistique, Paris 1882. S = E88ai de Semantique, Paris 1897. Brugmann Es = K. Brugmann, Uraprung des Scheinsubjekta '6a,' Leipzig 1914. KG = Kurze Vergleichende Grammatik, Strassburg 1904. VG = Grundris8 der Vergleichenden Grammatik, 2te Ausg., Strassburg 1897 ff. Versch = Verachiedenheiten der SatzgtBtaltung, Leipzig 1918. Brunot PL = F. Brunot, La Pen see et la Langue, Paris 1922. ChE = O. Jespersen, Chaptera on English, London 1918. Curme GG = G. O. Curme, A Grammar of the German Language, 2nd ed., New York 1922. Dan. = Danish. Delbriick GNS = B. Delbriick, Grundlagen de, Neuhochdeutachen Satzlehr., Berlin 1920. Synt. == Vergleichende Syntax der Indogermanischen Sprachen, Strassburg 1893. Deutachbein SNS = M. Deutechbein, System der Neuenglischen Syntaz, cathen 1917. Diez GRS = F. Diez, Grammatik der Romanischen Sprachen, 4ta Aufi., Bonn 1876. E. = English. Eliot FG = C. N. E. Eliot, A Finnish Grammar, Oxford 1890. ESt = Englische Studien. Falk & Torp DNS = Hjalmar Fo.lk og Alf Torp, Dan8k·noTaken, rynta1;, Kristiania 1900. Fr. = French. G. = German. Gabelentz Spr = G. T. d. Gabelentz, Die Spraehtoi88tMchajt, Leipzig 1891. Ginneken LP = J. v. Ginneken, Principes do Linguistique Pryckologique, Amsterdam, Pari. 1907. 11

12

THE PHII.. OSOPHY OF GRAMMAR

Gr. = Greek. Jespersen, (howth and Structure 0/ the E"I,gZish Language, 4J,h ed., Leipzig and Oxford 1923. Hanssen Sp. Gr. = F. Hanssen, Spani6Che Grammatik, Halle ) 910. IF = I1f,dogermanische Forschungen. It. = Italian. Keynes FL == J. N. Keynes, Studiu and Ezercia., in Formal Logic, 4th ed., London 1906. KZ = Kuhn's Zeitschri/t fur Vergleichmde Sprach/or,chung. Lang. (Language) = O. Jespersen, Language, its Nature, Development and Origin, London 1922. Lat. = Latin. LPh = O. Jespersen, Lehrbuch der Phonetik, 3te Aufl., Leipzig 1920. Madvig KI = J. N. MA.dvig, Kleine Philologi8che Sehri/ten, Leipzig 1875. MEG = O. Jcsp~rsen, Modern English G1'ammar, Heidelberg 1909, 1914. Meillet Gr = A. Meillet, Aperpu d'une Histoire de Ia Langue Grecque, Paris 1913. L! = Introduction () "etude du Languu Indo·Europtennu, 28 ed., Paris 1908. LH = Linguistique Historique et Linguistique Generale, Paris 1921. Meyer·Lubke Einfuhr. = W. Meyer-Lubke, Ein/uhrung in das Studium der Romanischen Sprachwissenschajt, 2te Aufl., Heidelberg 190!). Mikkelsen DO = Kr. Mikkelsen, Dan8k Ordfojningalcere, Kebenhavn 1911. Misteli = F. Misteli, Oharacieristik der hauptl/acht. Typen des Sprachbau8, Berlin 1892. MSL = M emoiru de 1a Societe de LinguiBtique. Fr. Muller Gr. = Friedrich Muller, Grundris. der Sprachwi9sen8chajt, Wien 1876. NED = A New English Dictionary, by Murray, etc., Oxford 1884 ti. Negation = O. Jespersen, Negation in English and Other Languages, Kebonhavn 1917 (Videnskabemes selskab, Hest). Noreen VS = A. Noreen, Vdrt Sprdk, Lund 1903 ft. Nygaard NS = M. Nygaard, Norren Syntax, Kristiania 1906. Nyrop Gr. = Kr. Nyrop, GrammairB Hi8tOriquIJ dB III Langu. Fra""aisll, Copenhague 1914 iI. OE = Old English. OFr. = Old French. OHG == Old High German. ON = Old Norse. Onions AS = C. T. Onions, An Advanced EngUsh Syntaz, London 1904. Paul Gr = H. Paul, Deutsche Grammatik, HoJle 1916 ft. P = Prinzipien der Sprachgeschichte, 7te Aufl., Halle 1909. PBB = (Paul und Bra.une), Beitriige zur Geachichte der deutachen Sprache. Pedersen GKS = H. Pedersen, Vergl. Grammatik der ke~ri8chen Sprachen. Gottingen 1909. RG == RUBsisk Grammatik. Kebenhavn 1916. PhG ... O. Jespersen, Phonstiache Grundjragen, Leipzig 1904. Poutsma Gr H. Poutsma, A Grammar oJ Late Modem English, Groningen 1904 ff.

as == O.

==

ABBREVIATIONS OF TITLES

13

Schleicher NV = A. Schleicher. Nomen uncl Verbum. Leipzig 1865. Schuchardt Br = Hugo Schuchardt-Brevier. v. L. Spitzer. Halle 1922. Sh. = Shakespeare. Sheffield GTh = A. D. Sheffield. Grammar ancl Thinking. New York 1912. Simonyi US = S. Simonyi. Die Ungari8che Sprache. Strassburg 1907. Sonnenschein = E. A. Sonnenschein. A New EngliBh Grammar. Oxford 1921 f. Sp. = Spanish. Spr. L. = O. Jespersen. SprogetB LogUc. K0benhavn 1913. Steinthal Charakteristik = H. Stein thai, CharakteriBtik der hauptBiichl. Typen deB SprachbautB. Berlin 1860. Stout AP =G. F. Stout, Analytic P8]Jchology. London lQ02. Streitoorg GE = W. Streitberg. Gotisches Elementarbuch,5te Aufi., Heidelberg 1920. Sweet CP = H. Sweet, Collected Papers, Oxford 1913. NEG = A New English Grammar. Oxford 1892, 1898. Tegner G = E. Tegner. Om Genus, Svenskan, Stockholm 1892. TG = The Terminology oj Grammar, by the Joint Committee (1911). 7th impr. 1922. Tobler VG = A. Tobler, VermiBchte Beitrage zur Franzo8ischm Grammatik. 3te Aufi., Leipzig 1921. US = United States. Vendryes L =J. Vendryes, Le Langage, Paris 1921. Vg. = Vulgar. Vondrak SG = W. Vondrak, Vergleichende Slavi8che Grammatik. Gl'lttingen 1906. Wa.ckemagel VS = J. Wa.ckernagel, Vorlesungen uber Synta:&, Baeel 1920. Wegener U = l)h. W('gener, Untersuchungen uber die Grundfragen des Spracl.lebens, Hallo 1885. Western R = A. Western, Norsk RikBl'I1dls-urammatikk, Kristiania 1921. Wilmanns DG = W. Wilmanns. Deutsche Grammatik, StraeBburg 1897 ft. Wundt S = W. WWldt, Die Sprache, Leipzig 1900.

PHONETIC SYMBOLS , stande before ~he stressed syllable. • indicaies length of 'he preceding sound. [a'] as in alms. [y] as in Fr. vu. [ai) as in ice. [A] as in cut. [au] as in house. ["} 8!1 in Fr. feu. [1El] as in hat. [re] as in Fr. Bwu.r. [-] French nasa.lir;ation. [eiJ as in hate. [e] lIB in care; Fr. tel. [e) as in G. ieh. [3] indistinct vowels. [x] as in G.. Se. loe1&. [i] as in hll; Fr. qui. [15] as in this. ri'] as in feel; Fr. hlle. [j] &iii in you. [oJ !18 in Fr. Beau. (J?] as in thick. [ouJ as in BO. U] as in .he. [0] open o·sound•. Ls] as in meaaure. [u) as in full; Fr. fON. ['] in Russian palatalization, in Dw:WIh glottal stop. [u'l &SI in fo~; Fr. epoU$$.

THE PHILOSOPHY OF GRAMMAR CHAPTER I

LIVING GRAMMAR Speaker and Hearer. Fonnulas and Free Exprel8iollll. Building up of Sentences.

Gramma1;ical Typea.

Speaker and Hearer. THE essence of language is human activity-activity on the part of one individual to make himself understood by another, and activity on the part of that other to understand what wa.s in the mind of the first. These two individuals, the producer and the recipient of language, or as we may more conveniently call them, the speaker and the hearer, and their relations to one another, should never be lost sight of if we want to understand the nature of language and of that part of language which is dealt with in grammar. But in former times this was often overlooked, and words and forms were often treated as if they were things or natural objects with an existence of their own-a conception which may have been to a great extent fostered through a too exclusive preoccupation with written or printed words, but which is fundamentally false, as will easily be seen with a little reBenon. If the two individuals, the producer and the recipient of language, a.re here spoken of as the speaker a.nd the hearer respectively, this is in consideration of the fact that the spoken and heard word is the primary form for language, and of far greater importance than the secondary form used in writing (printing) a.nd reading. This is evidently true for the countless ages in which mankind had not yet invented the art of writing or made only a sparing use of it; hut even in our modem newspaper-ridden communities, the vast majority of us speak infinitely more than we write. At any rate we shall never be able to understand what language is and how it develops if we do not continually take into consideration first and foremost the activity of speaking and hearing, and if we forget for a. moment that writing is only a. substitute for speaking. I 17

18

LIVING GRAMMAR

A written word is mummified until someone imparts life to it by transposing it mentally into the corresponding spoken word. The grammarian must be ever on his guard to avoid the pitfalls into whioh the ordinary spelling is apt to lead him. Let me give 8. few very elementary instances. The ending for the plural of substantives and for the third person singular of the present tense of verbs is in writing the same -8 in such words as endB, locks, riSel!, but in reality we have three different endings, as seen when we transcribe them phonetically [endz, loks, raiziz]. Similarly the written ending oed covers three different spoken endings in 8aaed, locked, ended, phonetically [seild, lokt, endid]. In the written language it looks as if the preterits paid and 8aid were formed in the same way, but differently from stayed, but in reality paid and stayed are formed regularly [peid, steid], whereas said is irregular as having its vowel shortened [sed]. Where the written language recognizes only one word there, the spoken language distinguishes two both as to sound and signification (and grammatical import), as seen in the sentence" There [~a] were many people there [1~E·a]." Quantity, stress, and intonation, which are very inadequately, if at all, indicated in the usual spelling, play important parts in the grammar of the spoken language, and thus we are in many ways reminded of the important truth that grammar should deal in the first instance with sounds and only secondarily with letters.

Formulas and Free Expressions. If after these preliminary remarks we turn our attention to the psychological side of linguistic activity, it will be well at once to mention the important distinction between formulas or formular units and free expressions. Some things in language-in any language--are of the formula character; that is to say, no one can change anything in them. A phrase like " How do you do 1 " is entirely different from such a phrase as " I gave the boy a lump of sugar. " In the former everything is fixed: you cannot even change the stress, saying " How do you do 1 " or make a pause between the words, and it is not usual nowadays as in former times to say "How does your father do 1 " or "How did you do 1" Even though it may still be possible, after saying" How do you do 1 " in the usual way to some of the people present, to alter the stress and say " And how do you do, little lVIary 1 " the phrase is for all practical purposes one unchanged and unchangeable formula. It is the same with" Good morning! ", "Thank you," "Beg your pardon," and other similar expressions. One may indeed analyze suoh a formula and show that it consists of several words. but it is felt and handled as a unit, whioh may often mean something quite

FORMULAS AND FREE EXPRESSIONS

19

different from the meaning of the component words taken separately; "beg your pardon," for instance, often means "please repeat what you said, I did not catch it exactly"; "how do you do' " is no longer a question requiring an answer, etc. It is easy to see that" I gave the boy a lump of sugar" is of a totally different order. Here it is possible to stress any of the essential words and to make a pause, for instance, after " boy," or to substitute" he " or " she" for" I," " lent" for" gave," " Tom" for "the boy," etc. One may insert "never" and make other alterations. While in handling formulas memory, or the repetition of what one has once learned, is everything, free expressions involve another kind of mental activity; they have to be created in each case anew by the speaker, who inserts the words that fit the particular situation. The sentence he thus creates may, or may not, be different in some one or more respects from anything he has ever heard or uttered before; that is of no importance for our inquiry. What is essential is that in pronouncing it he conforms to a certain pattern. No matter what words he inserts, he builds up the sentence in the same way, and even without any special grammatical training we feel that the two sentences John gave Mary the apple, My uncle lent the joiner five shillings, are analogous, that is, they are made after the same pattern. In both we have the same type. The words that make up the sentences are variable, but the type is fixed. Now, how do such types come into existence in the mind of a speaker ~ An infant is not taught the grammatical rule that the subject is to be placed first, or that the indirect object regularly precedes the direct object; and yet, without any grammatical instruction, from innumerable sentences heard and understood he will abstract some notion of their structure which is definite enough to guide him in framing sentences of his own, though it is difficult or impOBBible to state what that notion is except by means of technical terms like subject, verb, etc. And when the child is heard to use a sentence correctly constructed according to some definite type, neither he nor his hearers are able to tell whether it is something new he has created himself or simply a sentence which he has he&rd before in exactly the same shape. The only thing that matters is that he is understood, and this he will be if his sentence s in accordance with the speech habits of the community in which he happens to be living. Had he been a French child, he would have heard an infinite number of sentences like Pierre donne une pomme a. Jean, Louise a. donne sa. poupee a 8& soour, etc.,

20

LIVING GRAMMAR

and he would thus have been prepared to sa.y, when occasion arose, something like II va donner un sou a ce pauvre enfant.

And had he been a German boy, he would have constructed the corresponding sentences according to another type still, with dem and der instead of the French a, etc. (Of. Language, Ch. VII.) If, then, free expressions are defined as expressions created on the spur of the moment after a certain type which has come into existence in the speaker's subconsciousness as a result of his having heard many sentences possessing some trait or traits in common, it follows that the distinction between them and formulas cannot always be discovered except through a fairly close analysis; to the hearer the two stand at first on the same footing, and accordingly formulas can and do playa great part in the formation of types in the minds of speakers, the more so as many of them are of very frequent occurrence. Let us take a few more examples. " Long live the King!" Is this a formula or a free expression' It is impossible to frame an indefinite number of other sentenccs on the same pattern. Combinations such as "Late die the King! " or "Soon come the train I" are not used nowadays to express a wish. On the other hand, we may say" Long live the Queen" or" the President" or "Mr. Johnson." In other words, the type, in which an adverb is placed first, then a subjunctive, and lastly a subject, the whole being the expression of a wish, has totally gone out of the language as a living force. But those phrases which can still be used are a survival of that type, and the sentence" Long live the King" must therefore be analyzed as consisting of a formula" Long live," which is living though the type is dead, a subject which is variable. We accordingly have a sentence type whose use is much more restricted in our own days than it was in older English. In a paper on ethics by J. Royce I find the principle laid down "Loyal is that loyaIIy does." This is at once felt as unnatural, as the author has taken as a pattern the proverb" Handsome is that handsome does" without any regard to the fact that whatever it was at the time when the sentence was first framed, it is now to all intents and purposes nothing but a formula, a.a shown by the use of that without any antecedent and by the wordorder. The distinction between formulas and free expressions pervades all parts of grammar. Take morphology or accidence: here we have the sa.me distinction with regard to flexione.I forms. The plural eyen was going out of use in the sixteenth century; now the form is dead, but once not only that word, but the type a.ocord.ing

+

FORMULAS AND FREE EXPRESSIONS

21

to which it was formed, were living elements of the English language. The only surviving instance of a plural formed through the addition of -en to the singular is oxen, which is living as a formula, though its type is extinct. Meanwhile, ilwen, Jone, eyen, kine have been supplanted by 8hoM, JOeB, eyes, cows; that is, the plural of these words has been reshaped in accordance with the living t.ype found in kings, lineB, 8tones, etc. This type is now so universal that all new words have to conform to it: bicycleB, photos, kodn,1c8, aeroplanes, hooligans, ions, stunts, etc. When eyes was first uttered instead of eyen, it was an analogical formation on the type of the numerous words which already had -8 in the plural. But now when a child says eye8 for the first time, it is impossible to decide whether he is reproducing a plural form already heard, or whether he has learned only the singular eye and then has himself added -s (phonetically [z]) in accordance with the type he has deduced from numerous similar words. The result in either case would be the same. If it were not the fact that the result of the individual's free combination of existing elements is in the vast majority of instances identical with the traditional form, the life of any language would be hampered; a language would be a difficult thing to handle if its speakers had the burden imposed on them of remembering every little item separately. It will be seen that in morphology what was above called a " type" is the same thing as the principle of what are generally called regular formations, while irregular forms are" formulas." In the theory of word-formation it is customary to distinguish between productive and unproductive suffixes. An example of a productive suffix is -neBS, because it is possible to form new words like weariness, closeneB8, perverseness, etc. On the contrary -lock in wedlock is unproductive, and so is -th in width, breadth, heaUh, for Ruskin's attempt to construct a word illth on the analogy of weaUh has met with no success, and no other word with this ending seems to have come into existence for several hundred years. This is a further application of what we said above: the type adjective -neB8 is still living, while wedlock and the words mentioned in -th are now formulas of a type now extinct. But when the word width originated, the type was alive. At that far-off time it was possible to add the ending, which was then something like -~)t£, to any adjective. In course of time, however, the ending dwindled down to the simple sound l'(th), while the vowel of the first syllable was modified, with the consequence that the suffix ceased to be productive, because it was impossible for an ordinary man, who was not trained in historical grammar, to see that the pairs long: kngth, broad ; breadth, wide : width, deep : depth, whok : heaUh, dear: dearth. represented one and the Same type of formation. These words

+

22

LIVING GRAMMAR

were, accordingly, handed down traditionally from generation to generation as units, that is, formulas, and when the want was felt for a new ' abstract noun' (I use here provisionally the ordinary term for such words), it was no longer the ending -th that was resorted to, but -ne88, because that offered no difficulty, the adjective entering unchanged into the combination. With regard to compounds, similar considerations hold good. Take three old compounds of hus 'house,' hUsbande, hUsJ>ing, huswif. These were formed according to the usual type found in innumerable old compounds; the first framers of them conformed to the usual rules, and t!lUS they were at first free expressions. But they were handed down as whole, indivisible words from generation to generation, and accordingly underwent the usual sound changes; the long vowel u was shortened, [s] became voiced [z] before voiced sounds, [1'] became [t] after [s], [w] and [f] disappeared, and the vowels of the latter element were obscured, the result being our present forms husband, husting(s), hussy, phonetically [hAzband, hAsti7]z, hAzi]. The tie, which at first was strong between these words and hus, was gradually loosened, the more so because the long 'It had here become a diphthong, house. And if there was a divergence in form, there was as great a divergence in meaning, the result being that no one except the student of etymology would ever dream of connecting husband, hustings, or hussy with house. From the standpoint of the living speech of our own days the three words are not compound words; they have, in the terminology here employed, become formulas and are on a par with other disyllabio words of obscure or forgotten origin, such as 80pha or cousin. With regard to huswif there are, however, different degrees of isolation from house and wife. Hussy [hAzi] in the sense • bad woman' has lost all connexion with both; but for the obsolete sense' needle-case' old dictionaries record various forms showing oonflicting tendencies: huswife [hAzwaif], hussif [hAzi£], kUS8ive; and then we have, in the sense of 'manager of a house,' hnusewife, in which the form of both components is intaot, but this appears to be a oomparatively recent re-formation, not recognized, for instanoe, by Elphinston in 1765. Thus the tendenoy to make the old oompound into a formula was oounteracted more or less by the actual speech-instinot, which in some applications treated it as a free expression: in other words, people would go on combining the two elements without regard to the existence of the formular oompounds, wh!ch had become more or less petrified in Bound and in menning. This phenomenon is far from rare: grindstone as a formula had beoome [grinstan] with the usual shortening of the vowel in both elements, but the result of a free combination has prevailed in the current pronunoiation [gra.ind-

FORMULAS AND FREE EXPRESSIONS

23

stoun]; in waistcoat the new [weistkout] is beginning to be used instead of the formular [weskJt]; fearful is given as sounding 'ferful' by eighteenth-century orthoepists, but is now always [fi,)£(u)I]. For other examples see MEG I, 4. 34 ff. Something similar is seen in words that are not compounds. In Middle English we find short vowels in many comparatives: deppre, grettre as against deep, great (greet). Some of these compara.tives became formulas and were handed down as such to new generations, the only surviving instances being latter a.nd utter, which have preserved the short vowels because they were isolated from the positives late and out and acquired a somewhat modified meaning. But other comparatives were re-formed as free oombinations, thus deeper, greater, and in the same way we have now later and outer, which are more intimately connected with late and 01IJ than latter and utter are. Stress presents analogous phenomena. Chlldren, of course, learn the accentuation as well as the sounds of each word: the whole of the pronunciation of a word is in so far a formular unit. But in some words there may be a conflict between two modes of accentuation, because words may in some instances be formed as free expressions by the speaker at the moment he wants them. Adjectives in -able, -ible as a rule have the stress on the fourth syllable from the ending in consequence of the rhythmic principle that the vowel which is separated by one (weak) syllable from the original stress is now stressed, thus 'despicable 1 (originally as in French ,dcspi'cable), 'comparable, Ilamentable, 'preferable, etc. In some of these the rhythmic principle places the stress on the same syllable as in the corresponding verb: con'siderable, 'violable. But in others this is not so, and a free formation, in which the speaker was thinking of the verb and then would add -abk, would lead to a different accentuation: the adjective corresponding to aclcept was 'acceptabk in Shakespeare and some other poets, and this formula still survives in the reading of the Prayer Book, but otherwise it now is reshaped as ac1ceptable: refutable was [Ire£jutabl], but now it is more usual to say [rilfju'wbl]; 'respectabk has given way to relspectabZe;, Sha.kespeare's and Spencer's Idetestabk has been supplanted by de1tcstabk, which is Milton's form; in admirabk the new [ad1mairabl] has been less successful in lupplanting [laldmirabl], but in a great many adjectives analogy, i.e. free formation, has prevailed entirely: algreeable, de'plorabk, re1marlcabk, irrelsistibk. In words with other endings we have the same conflict: lconfcssor and conljcs8or, ca'pitali8t and Icapitali8t, de1momtraI Full stress il here indioated by a .hon vertioal stroke above, and half• • tresa by a shon vertioal stroke below-theM marka placed before the begin. ning of the streaaed syllable in aooOl'danoe with the practice now followed by moat phonetioilPdll.

LIVING GRAMMAR

24

tive and 'demonstrative, etc., sometimes with changes of meaning, the free formation following not only the accent, but also the signification of the word from which it is derived, while the formula has been more or less isolated. (Examples see MEG Ch. V.) The British advertiBement [ad1va·tizmant] shows the traditional formula, the American pronunciation [1m